Commits · f1c2370c48cd00ae0f35cc2dcd10c762414792d1 · Roger Ferrer / llvm-epi · GitLab

Nov 19, 2015
- Expand subregisters in MachineFrameInfo::getPristineRegs · df537b97
  Krzysztof Parzyszek authored Nov 19, 2015
  df537b97
- Fix the debug build breakage that getDiscriminator is called by mistake. · 014fb557
  Dehao Chen authored Nov 19, 2015
  014fb557
- Revert r253253 and r253126: "Don't recompute LCSSA after loop-unrolling when possible." · 6c11c04d
  Michael Zolotukhin authored Nov 19, 2015
  6c11c04d
- Reimplement discriminator assignment algorithm. · 23e2278e
  Dehao Chen authored Nov 19, 2015
  23e2278e
- [AddressSanitizer] assert(false) -> llvm_unreachable and remove return. · 193c4edf
  Davide Italiano authored Nov 19, 2015
  193c4edf
- [AArch64] Refactoring aarch64-ldst-opt. NCF. · c12c2790
  Jun Bum Lim authored Nov 19, 2015
  c12c2790
- [LIR] Fix 80-column from previous commit. · 3ecc8d8d
  Chad Rosier authored Nov 19, 2015
  3ecc8d8d
- [LIR] Sink checks into function to enable future refactoring. NFC. · fddc01f3
  Chad Rosier authored Nov 19, 2015
  fddc01f3
- [GlobalOpt] Localize some globals that have non-instruction users · 1d695a09
  James Molloy authored Nov 19, 2015
  1d695a09
- update comment and error message; NFC · 2fe77282
  Sanjay Patel authored Nov 19, 2015
  2fe77282
- [LIR] Use the more appropriate method. NFC. · 85c21f0a
  Chad Rosier authored Nov 19, 2015
  85c21f0a
- [AArch64]Extend merging narrow loads into a wider load · 4c35ccac
  Jun Bum Lim authored Nov 19, 2015
  4c35ccac
- [CGP] despeculate expensive cttz/ctlz intrinsics · 4699b8ab
  Sanjay Patel authored Nov 19, 2015
  4699b8ab
- X86: More efficient legalization of wide integer compares · dcc25004
  Hans Wennborg authored Nov 19, 2015
  dcc25004
- TargetParser.cpp: Fixup -- StringRef::startswith() is better here. NFC. · 768579c4
  NAKAMURA Takumi authored Nov 19, 2015
  768579c4
- SamplePGO - Sort samples by source location when emitting as text. · ef548d29
  Diego Novillo authored Nov 19, 2015
  ef548d29
- llvm/lib/Support/TargetParser.cpp: Rework llvm::ARM::getArchExtFeature() to... · b6b25458
  NAKAMURA Takumi authored Nov 19, 2015
  b6b25458
- [LV] Add a helper function, isReductionVariable. NFC. · 33efdf81
  Chad Rosier authored Nov 19, 2015
  33efdf81
- [mips] Expansion of ROL and ROR macros · 00f998b4
  Zoran Jovanovic authored Nov 19, 2015
  00f998b4
- AVX-512: Fixed COPY_TO_REGCLASS for mask registers · 7c2c9fd2
  Elena Demikhovsky authored Nov 19, 2015
  7c2c9fd2
- [X86][AVX] Fix lowering of X86ISD::VZEXT_MOVL for 128-bit -> 256-bit extension · 846b64e1
  Simon Pilgrim authored Nov 19, 2015
  846b64e1
- Alternative to long nops for X86 CPUs, by Andrey Turetsky · b7b82bf3
  Alexey Bataev authored Nov 19, 2015
  b7b82bf3
- [FunctionAttrs] Provide a mechanism for adding function attributes from the command line · 0ecdbe7d
  James Molloy authored Nov 19, 2015
  0ecdbe7d
- AVX512: Implemented encoding, intrinsics and DAG lowering for VMOVDDUP instructions. · 1f782968
  Igor Breger authored Nov 19, 2015
  1f782968
- AVX512: Implemented encoding for the vmovss.s and vmovsd.s instructions. · 4424aaa2
  Igor Breger authored Nov 19, 2015
  4424aaa2
- AVX512: Implemented encoding for the follow instructions. · 81b79de5
  Igor Breger authored Nov 19, 2015
  81b79de5
- Pointers in Masked Load, Store, Gather, Scatter intrinsics · 1ca72e18
  Elena Demikhovsky authored Nov 19, 2015
  1ca72e18
- Revert "Change memcpy/memset/memmove to have dest and source alignments." · 67cf9a72
  Pete Cooper authored Nov 19, 2015
  67cf9a72
- Do not require a Context to extract the FunctionIndex from Bitcode (NFC) · 354f520f
  Mehdi Amini authored Nov 19, 2015
  354f520f
- Fix bug 25440: GVN assertion after coercing loads · b69babd0
  Weiming Zhao authored Nov 19, 2015
  b69babd0
- [X86] Enable shrink-wrapping by default. · 46d5c711
  Quentin Colombet authored Nov 19, 2015
  46d5c711
- Fix several long lines (>80) in LoopVectorize.cpp. NFC. · 7b2ae9ab
  Cong Hou authored Nov 19, 2015
  7b2ae9ab
- [SimplifyLibCalls] New trick: pow(x, 0.5) -> sqrt(x) under -ffast-math. · c5cedd19
  Davide Italiano authored Nov 18, 2015
  c5cedd19
- [AArch64] Enable shrink-wrapping by default. · f6645cce
  Quentin Colombet authored Nov 18, 2015
  f6645cce
Nov 18, 2015
- Fix returned value for GVN: could return "false" even after modifying the IR · adb4057a
  Mehdi Amini authored Nov 18, 2015
  adb4057a
- Minor cleanups (from review feedback) · cfb14565
  Xinliang David Li authored Nov 18, 2015
  cfb14565
- [BuildLibCalls] EmitStrNLen() is dead code. Garbage collect. · 455ea11d
  Davide Italiano authored Nov 18, 2015
  455ea11d
- Change memcpy/memset/memmove to have dest and source alignments. · 72bc23ef
  Pete Cooper authored Nov 18, 2015
  72bc23ef
- [DAGCombiner] Vector constant folding for comparisons · c1a46b72
  Simon Pilgrim authored Nov 18, 2015
  c1a46b72
- ARM: make sure backend is consistent about exception handling method. · 747ae9a7
  Tim Northover authored Nov 18, 2015
  747ae9a7